ABSTRACT: In a minority of diabetic patients, glycaemic control can be difficult to achieve. This may lead to persistent clinical signs, recurrent hypoglycaemia or fluctuating insulin requirements. The causes of poor control can be broadly divided into difficulties with the diabetic regimen or concurrent illness in the patient. Successful stabilisation requires knowledge of the possible problems that can underlie instability and a thorough and logical approach to the investigation of these potential causes.

A difficult diabetic is one which is hard to stabilise. This can occur during the initial stabilisation period or an animal may develop problems when it had been previously well controlled. The majority of diabetic cats and dogs can be stabilised without difficulty. Those that cannot, usually fall into three categories with respect to clinical presentation:

   persistent clinical signs in spite of adequate therapy

   recurrent hypoglycaemia

   fluctuating insulin requirements.

What is good control?

A well-controlled diabetic should have resolution of clinical signs. There should be no polyuria, polydipsia or polyphagia. As well as this, the animal should have no hypoglycaemic episodes and it should maintain a normal body weight.

Expected insulin doses are approximately 1.0-1.5 IU/kg per dose for once daily therapy and up to 1.0 IU/kg per dose for twice daily therapy.

Blood tests are a very useful adjunct to clinical signs for monitoring diabetics and guidelines for expected values are given in Table 1. It is important to use these in combination with clinical signs and never to adjust insulin doses on the basis of laboratory results alone.

Causes of poor stability

The main reasons for instability are:

1.   Management problems, including an incorrect routine or problems with the insulin; such as incorrect administration, poor absorption, spoilt or out-of-date product.

2.   Short duration of action of insulin (Figure 1)

Figure 1: Glucose curve demonstrating short duration of action of insulin.

This dog was on once daily injections of lente insulin given at time 0 with a meal. The red lines indicate the maximum and minimum acceptable levels for blood glucose throughout the day

3.   Insulin-induced hyperglycaemia (Somogyi overswing), which is where there is a dramatic fluctuation in the blood glucose throughout the day – from very low to very high (Figure 2).

Figure 2: Glucose curve demonstrating insulin-induced hyperglycaemia or Somogyi overswing. This dog was on once daily injections of lente insulin given at time 0 with a meal

4.   Insulin resistance – this is where another condition reduces an animal’s response to insulin. (Figure 3).

Figure 3: Glucose curve demonstrating insulin resistance. This cat was injected with PZI insulin twice daily at times 0 and 12 hours, with a meal at each dose

A problem in the first category is generally picked up on history taking, whereas the last three are usually identified on a  glucose curve. A glucose curve is a graph showing the blood glucose concentration throughout the day.


Investigation to determine the cause of the instability should follow a thorough, logical approach (Table 2).

The first step is to take a careful history. This should include the following questions:

   Have there been any changes to the routine?

   Are the dose, frequency and timings of insulin injections the same?

   Are the syringes the correct ones for the insulin prescribed – 40IU/ml or 100 IU/ml? This is especially important if owners are purchasing supplies outside the practice.

   Is the insulin in date? Has it been diluted?

   Are any treats being fed?

   Has the vial been refrigerated? Could it have been frozen inadvertently?

   Has it been mixed gently?

   Is the diet still the same?

   Is the exercise regular in the case of a dog?

   Have other medications been given?

   Has a bitch been spayed?

It is important to evaluate the owner’s diabetic diary and ask questions. Ensure any dose changes were appropriate.

The dose should be increased by no more that 10 per cent each time, and no more frequently than every two to three days. More rapid changes can cause a Somogyi overswing.

The next step is to perform a full clinical examination and weigh the patient. It might then be necessary to watch the owner inject the animal, unless the pet was previously well controlled. It is important to approach this in a tactful manner, as some owners may be affronted if they think you are questioning their competence. Also enquire whether one owner has been giving all the injections.

When observing, check that they are gently mixing the vial, drawing up the correct dose without air bubbles; that their injection technique is correct; and that they are not injecting into hair by mistake. Also check the injection site for any reaction which could affect absorption.

It is best to confirm that control is not adequate by testing blood glucose, ideally at the nadir and possibly pre-insulin injection too and measuring serum fructosamine (Table 1).

If the cause of the instability is not apparent by this stage, then a glucose curve is required. The blood glucose is measured every one to two hours for 12 – 24 hours and the values plotted on a graph wit
h time on the x axis and glucose concentration on the y axis. The curve should be continued for 12 hours in animals on twice daily therapy and 24 hours for those on once daily therapy.

Remember, if you perform a curve assessment on two consecutive days, you’ll see different results! It is to be used as a guide. Continuous glucose monitors, such as the iPRO™ (Medtronic), are available now for cases where more frequent measurement of glucose is required. At present, they are mainly a referral tool because of the associated costs, but their use in practice is becoming more widespread.

When evaluating a curve, there are certain features that should be assessed. These are:

   onset of action of insulin

   nadir glucose

   duration of action of insulin

   Is there hyperglycaemia?

   Is there Hypoglycaemia?

   When are the ideal meal times?

Figures 1,2 and 3 show the three main patterns detected on a glucose curve.

In Figure 1, the insulin is not lasting long enough – sometimes referred to as rapid metabolism of insulin. The solution is to increase the frequency of the injections to twice daily. It is now thought the majority of dogs and all cats require twice daily therapy with lente insulin preparations.

In Figure 2, there is marked hypoglycaemia, caused by the insulin. The animal then responds to this by producing hormones to antagonise insulin, which leads to rebound hyperglycaemia. This is called a Somogyi overswing. The action required is to reduce the dose of insulin by a third or to a starting dose, to prevent the hypoglycaemia occurring.

In Figure 3, there is no real reduction in blood glucose throughout the day. This is termed insulin resistance, and should be considered if the dose of insulin being administered is high – for instance, > 2.2 IU/kg per dose for once daily therapy and >1.5 IU/kg per dose for twice daily therapy.

This pattern will also be seen in animals that are on a dose of insulin that is too low, which is a common cause of suspected instability.

Insulin resistance generally occurs because of concurrent disease states. If this pattern is seen, the dose should be increased to attain best stability whilst looking for the underlying cause. There are numerous potential underlying causes including:

   dioestrus diabetes in unspayed bitches

   infection, such as urinary tract, dental disease

   steroid or progestagen therapy


   acromegaly (cats)

   obesity (Figure 4)

Figure 4: Obesity can lead to insulin resistance. Some obese diabetic cats may experience resolution of their diabetes once they have lost sufficient weight

   poor absorption


   hyperthyroidism (cats)

   hypothyroidism (dogs)

   anti-insulin antibodies

   renal failure


If the underlying cause waxes and wanes, this can lead to fluctuating insulin requirements. It is beyond the scope of this article to look in detail at diagnosis of all these conditions, but if a pattern of insulin resistance is detected, likely causes should be ruled in or out systematically.

The final category of a difficult diabetic is recurrent hypoglycaemia. This can usually be indentified from clinical signs, low blood glucose measurements at both the nadir and pre-injection and also low serum fructosamine levels. It may be that the dose of insulin is too high but in cats it may also be the consequence of transient diabetes or diabetic remission.

Between 20 and 40 per cent of diabetic cats go into remission, usually in the first one to three months after diagnosis. Remission may last from weeks to years. If the cat is on a low dose of insulin – 1 unit per injection – and serum fructosamine and blood glucose are low, then the insulin should be discontinued and the cat monitored closely for recurrence of signs.


Many veterinary nurses run diabetes clinics in their practices. They are responsible for teaching owners of new diabetic patients how to inject their pets and helping them to establish the diabetic routine.

For problem cases – such as those discussed in this article – the VN can perform initial investigations, such as evaluating the owner’s diabetic diary with them, checking that they are happy and competent at injecting insulin, that there is no problem with the insulin, and testing the urine and blood glucose levels.

Further investigations usually involve a consultation with a veterinary surgeon, who will need to examine the animal. Then the VN could perform a glucose curve if this is indicated or help with any further tests that might be required. 


Jelena Ristic


Since qualifying from the Royal Veterinary College in 1992, Jelena Ristic has worked in the small animal field. This has included a mixture of general practice, internal medicine referral work and clinical pathology. She currently divides her time between small animal practice in Hertfordshire and working for Axiom Veterinary Laboratories. Jelena holds the RCVS Diploma in Small Animal Medicine.

To cite this article use either

DOI: 10.1111/j.2045-0648.2010.00004.x or

Veterinary Nursing Journal Vol 26 pp15-17


1.   HESS. R. S. and WARD, C. R. 120001 Effect of insulin dosage on gtycemic response in dogs with diabetes mellitus: 221 cases (1993-1998). JAVMA Jan 15;216121:217-21.

2.   BERTOY. E. H. NELSON. R. W. and FELDMAN.E. C. (1995) Effect of lente insulin for treatment of diabetes mellitus in 12 cats. JAVMA June 1 ;206 (111:1729-311729-31.

3.   Martin GJ. Rand JS. 120011 Pharmacology of a 40 lU/ml porcine lente insulin preparation in diabetic cats: findings during the first week and after 5 or 9 weeks of therapy. J Feline Med Siirg. Mar;3(1l:23-30.

Further reading

GRAHAM, P. A. I2004I The uncontrollable diabetic. In: BSAVA Ma
nual of Canine and feline endocrinology 3rd edn, ed CT Money and ME Peterson, pp 66-75.


Veterinary Nursing Journal • VOL 26 • January 2011 •